On Thu, May 23, 2013 at 6:10 PM, Jesse Barnes <jbarnes at virtuousgeek.org> wrote: >> - For the clock readout code I think we should be able to have pipe config >> compare support (with adjusted_mode->clock), with a bit of fuzz at >> least. Not on current dinq, but with my cleanup to give >> adjusted_mode->clock saner semantics: >> >> http://permalink.gmane.org/gmane.comp.freedesktop.xorg.drivers.intel/21750 > > So far I haven't needed the fuzz, but I'm not opposed to adding it as > needed. I mean for the pipe_config comparison/checking in general, which seems to still be lacking. >> - There's no reason imo for the new ->get_clock vfunc, it splits exactly >> the same as the existing ->get_pipe_config. So could simply be included >> there. > > Except that it depends on the results of the encoder->get_config > callback... that's the only reason I split it. If you have ideas about > how to merge it back that's ok with me. Yeah, I always forget about that. I still think we should shovel everything into the platform get_config (including the encoder loop) like we do for the crtc enable/disable functions. But the current approach still works, so meh ;-) -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch
